By Catherine Case (University of Georgia)
Information
In their book Beyond Constructivism (2003), Lesh and Doerr describe a “models and modeling perspective” that conceptualizes learning as model-building – an iterative process in which students invent, extend, and revise constructs. Central to the models and modeling perspective is the assumption that reality is accessed through models and representations that may emphasize different aspects of the underlying system. This poster explores how students interact with the models, tools, and representations of statistical inference and illustrates ways in which “model-eliciting activities often function as thought-revealing activities” (Lesh & Doerr, 2003, p. 31). Traditional inference methods use theoretical probability distributions (e.g., Normal distribution, t distribution, Χ^2 distribution) to model the outcomes that would occur by chance under the null hypothesis, and students use tools such as calculators or statistical software to interact with those models. Alternatively, simulation-based inference methods model chance outcomes using simulations, which employ physical chance devices (e.g., coins, dice, spinners) or a computer to mimic a random process. The poster will include examples of student work collected through task-based interviews of AP Statistics students familiar with both inference methods. Attendees will be invited to consider how use of particular models and tools make student thinking visible (or not!) in the interview examples and in their own classroom experience.